Download Plasma Clean to Reduce Wire Bond Failures

Plasma Clean to Reduce Wire
Bond Failures
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
Introduction
Many books and articles have been written by wire bonding experts about wire
bonding. Very often, plasma treatment is referred to therein as a means to influence
the bonding process or the long‐ term reliability of the bond. Significantly fewer,
however, are the articles written by plasma experts on the applications of plasma in
microelectronics packaging, and in particular, on the use of plasma prior to wire
bonding. One reason for this, in a world which is acclimatized to statistical process
control, may be that plasma deals with the “unknown”, that which is “out of control”. A
further reason may be that a plasma process is defined by quite a large number of
parameters, and it’s not always clear why one combination of settings works for one
application while a second application, to all intents and purposes analogous, works
best with a completely different set of parameters. Plasma has something of the aura
of “alchemy” or a “black box”. There are, nonetheless, realistic expectations about what
plasma can contribute both to the performance of a wire bonding process and to the
long‐term reliability of the packaged device. There is also a certain logic to be followed
in developing a plasma process, even if it is not so hard and fast as some may like.
The benefits of plasma are to be found in two different areas: the wire bonding process
itself, or what we might call “the statistics”, and the long‐term reliability of the device, or
what we shall refer to as “device reliability”. Although the two are to some extent
interrelated, it is easiest, for clarity’s sake, to treat them separately.
Plasma and the “Statistics” of the Wire Bonding Process
Variety of Bonding Surfaces
Today, most “first” bonds are still made to an aluminium metallization on a
semiconductor device and most “second” bonds to gold. This picture, however, is
changing very rapidly with the introduction of copper (and other) wires and of new
metallizations on both the devices and the leadframes, or substrates, they are bonded
to. Here, we already encounter a source of diversity in the plasma process: flash gold
© Nordson MARCH
Page 1
www.NordsonMARCH.com
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
of 50nm may be treated very differently than 2 microns of thick gold. A nickel/palladium
metallization on an IC will be treated differently than aluminium. Metal leadframes can
be treated very differently than BGA substrates.
While tight process control in the semiconductor industry does tend to result in a rather
predictable metallization, the processes in the printed circuit board and plating
industries inherently show more variability. It is not strange to source BGA substrates
from three vendors, all who were supplied the same specification, and find that they
result in three different sets of wire bond statistics.
Contamination on the Surface to be Bonded
In an ideal world we would be bonding onto clean metal surfaces, with the possible
exception of a thin oxide layer, on both the semiconductor device and the substrate or
leadframe we are connecting it to.
In practice, there are a number of sources of surface contamination which influence the
wire bond statistics and the device reliability.
These include:
•
•
•
•
•
Inorganics, particularly fluorine, on the IC pad whose origins lie in the wafer
processes which precede singulation.
Organic contamination arising from outgassing and bleeding of the die attach
adhesive. These can be found both on the IC metallization and on the
leadframe or substrate.
“Atmospheric” contamination which is present in the air and which deposits
onto the bond pads. This is largely organic, but traces of inorganics from the
atmosphere are also often present.
Products of “diffusion” resulting from grain boundary migration of underlying
metal layers: best known are nickel migration through flash gold and palladium
migration.
Excessive oxide on the surface of aluminium and copper.
Wire bond statistics tend to be influenced primarily by the presence of organics and
oxides while long term reliability is more generally influenced by inorganic
contamination. There is, however, an area of overlap.
© Nordson MARCH
Page 2
www.NordsonMARCH.com
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
While plasma can, in principle, deliver a perfectly clean metal surface for wire bonding,
it has been repeatedly demonstrated that a metal surface becomes contaminated with
atmospheric organics within just a few hours of plasma treatment. This effect is
particularly noticeable in a cleanroom which, although “particle free”, can have high
concentrations of organics arising from the plastics, surface finishes, and people
present and which can be concentrated through recirculation of cleanroom air.
Effects of Surface Contamination on Wire Bond Statistics
The problem with contaminants is simply that they cover the surface we want to bond
to, preventing us from making a good bond. In addition, the presence of contaminants
is “non‐systemic”; they appear and disappear, are non‐uniformly distributed and almost
impossible to locate, identify, and quantify. A bonding process can run for hours or
days without issues and suddenly go out of control for no apparent reason. The
immediate solution is often to turn up ultrasonic power (change of a controlled
parameter) with all the accompanying consequences.
The key issues we see are:
•
•
•
•
•
•
Non‐Stick on Pad (NSOP)
Lifts
Decrease in average shear strength
Decrease in bonded area
Decrease in average wire‐pull strength
Cratering and other damage arising from a too aggressive bond process
Lifts are regarded differently by different operators. For some operators a lift which is
within specification is regarded as acceptable. For most applications requiring some
degree of reliability a lift is never an acceptable failure mode. The question is whether
you will capture a few randomly distributed “lifts” which have not yet revealed
themselves as such. Decreases in bonded area, average shear strength, and pull
strength are all indicators that when the bond is made, the area of contact between the
bond wire and the pad will be sub‐optimal. This is due to the surface contaminant
impeding the process of “welding”. In addition, it is known that, at the moment the bond
is made, the contact area is never one hundred percent, but it increases with time to
give a stronger bond through the process of grain boundary migration. Surface
contamination will interfere with this process, leading to a bond which does not reach
© Nordson MARCH
Page 3
www.NordsonMARCH.com
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
its maximum potential strength. This is one area where sub‐optimal wire bond statistics
will influence device reliability. If the bond area is not maximized the bond will fail
earlier during reliability testing.
Influence of Plasma Treatment on Wire Bond Statistics
Turning the above around, it’s easy to deduce that the benefits of plasma include the
following:
•
•
•
•
•
•
•
Reduction or elimination of NSOPs
Reduction or elimination of lifts
Increase in average shear strength
Increase in bonded area
Increase in average wire‐pull strength
Increase in process window for the bonding process
Reduction or elimination of cratering and other damage arising from a too
aggressive bond process
Since as we mentioned, contaminants cover the bonding surface preventing a good
bond, it is quite possible that plasma will serve to eliminate short‐term, non‐systemic
excursions from a process that is otherwise running very stable with a high process
capability. However, even a highly capable process which is running below its ultimate
potential will deliver sub‐optimal device reliability. An optimized wire bond process
which maximizes bonded area and minimizes pad damage will always deliver better
reliability than a sub‐optimized process.
Plasma and Device Reliability
Failure Modes during Reliability Testing
Despite the considerable variety now found in combinations of bond wires and
metallizations, the failure modes which are seen during reliability testing tend to be
common to all metallurgical systems. Monometallic systems form a “class within a
class” in this case rather than an exception.
The difference from one metallurgical system to another is not so much the failure
mode as the susceptibility to this failure mode: the time to failure.
© Nordson MARCH
Page 4
www.NordsonMARCH.com
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
With some simplification we can say that the most predominant causes of failure
involve the following:
•
•
•
Poorly formed bonds fail much faster than well‐formed bonds,
Well‐formed bonds which are free of contaminants will ultimately fail by
Kirkendall voiding (polymetallic systems), but usually well beyond the required
lifetime of the package
Failures of “well‐made bonds” which compromise the reliability of the package
are usually the consequence of contaminant accelerated voiding (Horsting
Effect) and/or corrosion due to contaminants
Framed in this way, the connection with bond pad contamination is readily apparent.
Assuming that our bonding process is well optimized and that our metallizations are “in
order” in terms of thickness, adhesion, surface topography, density, and so on, the
cause of a poorly formed bond, i.e. a bond with a low bond area consisting largely of
non‐coalesced microbonds, is almost always organic contaminants on the surface of
the bond pad. The reasons why such bonds fail very much faster than we might expect
are somewhat complex and go beyond the scope of this overview.
However, the solution to the problem‐‐ “clean your bond pad” ‐‐ is delightfully simple.
Most common polymetallic systems form intermetallics. The most widely studied is
undoubtedly the gold‐aluminium combination. The formation of an intermetallic is the
essential first step of forming the bond and intermetallics will be formed during the
production process, burn‐in, and use of the components. Given that intermetallic
formation is essentially the diffusion of one metal into the other (with the creation of
chemical compounds between the two metals), and that eventually one of them will be
exhausted by this process, the end result is always failure. When the bond is clean of
contaminants, however, the “time to failure” is usually much longer than the design
lifetime of the part and so intermetallic formation and the resultant Kirkendall voiding is
not a problem in practice. The intermetallics are strong and electrically conductive.
The problems arise when the surfaces which are bonded are not clean, but are
contaminated with inorganic materials, in particular halogens. The halogens may be
present as a result of wafer processes, environmental contamination, or may be
present in the molding compound used in the device package. By a somewhat
complicated mechanism known as the Horsting Effect and, like Kirkendall voiding, a
consequence of the diffusion of one metal into the other, the halogens become
© Nordson MARCH
Page 5
www.NordsonMARCH.com
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
concentrated into zones at the metal interface where they greatly accelerate the
process of void formation, but also represent a site at which corrosion will take place.
Halogens which are not incorporated in the bond interface, but which are in contact
with the metal surface, particularly at the perimeter of micro‐welds, are able to form
electrochemical cells which lead to rapid corrosion. A poorly formed bond, with low
bond area and a large number of micro‐welds, will have a greater perimeter length
than a well‐formed bond and so be more susceptible to this corrosion mechanism.
Of the three failure mechanisms mentioned above, this latter is the most prevalent
since the majority of integrated circuits today make use of gold wire bonding onto an
aluminium metallization. Despite the complexity of the failure mechanism, the solution
is once again delightfully simple: “clean your bond pad”.
Figure 1. Bench‐top plasma cleaning
system for surface activation and
adhesion improvement (AP600)
Developing a Plasma Process
We have reached a point where it’s easy to believe that by ensuring a clean surface to
which we can wire bond plasma it will always result in improved “wire bond statistics”
(even if it’s only by eliminating non‐systemic excursions) and improved device
reliability. In our “ideal world” we would use a plasma which sputters all the organic and
inorganic contaminants from our bond pads and delivers the perfectly clean bond pad
leading to maximized “wire bond statistics” and maximized “device reliability”. This,
however, is where the complexity begins.
© Nordson MARCH
Page 6
www.NordsonMARCH.com
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
This solution, which would be a high power argon direct plasma at relatively low
pressure, is sometimes used. It can be used on some metal leadframes with power
devices. The issues that limit its applicability are the effects of sputtering and
overheating. Metal leadframes are not typically susceptible to overheating, and the
re‐deposition of sputtered material is only an issue when it is significant and results in a
change in surface resistivity or device performance, i.e. leakage currents or changes in
device characteristics. Sputtering away organic contamination, which we have seen
has the largest effect on wire bond statistics, is, however, much slower than removing
it in a “chemical plasma” such as oxygen, but oxygen plasma is almost ineffective
against inorganic contaminants. Oxygen is also not indicated in most instances where
we are dealing with oxidizable metals such as copper or palladium.
Here we have the first “polarity” that we need to consider when designing a plasma
process: sputtering plasma or chemical plasma? In most cases we choose a process
which combines both effects, i.e. mixtures of oxygen and argon. For oxidizable
surfaces this is most commonly a mixture of hydrogen and argon. This choice leads
directly to the next polarity. Chemical plasmas work best at higher pressures (250 –
2000mT) whereas sputtering plasmas require low pressures (150 – 250mT) to
maximize the mean free path of the energetic ions that perform the sputtering. In fact,
since organic contamination is our most common issue, we typically start with a higher
pressure oxygen plasma and tend to make it more aggressive (by lowering pressure,
adding argon, and increasing plasma power) if we see that we are dealing with
significant amounts of inorganic contamination or if there is an indication that we could
reduce cycle time, and therefore increase production throughput, by using a more
aggressive plasma process. Turning up the power too much, particularly with organic
substrates, can result in overheating and can result in excessive sputtering, for
example of flash gold. So depending on the parts we want to wire bond, leadframe or
organic substrate, thick or thin gold, sensitive or robust components, very fine pitch or
large pitch (at very fine pitch changes in surface resistivity become more important) we
manipulate the “levers” of the plasma process.
The objective is to define a process window where we maximize the cleaning effect,
but avoid the potential “downside” of an over‐aggressive plasma. Given the number of
variables involved, both in the nature of the parts and the plasma parameters that can
be varied, a Design of Experiment (DoE) is often a useful approach. One has to bear in
mind that DoEs, like SPC, are designed to work in a world of controlled and predictable
“cause and effect,” which contamination normally is not!
© Nordson MARCH
Page 7
www.NordsonMARCH.com
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
Going Into Production
From Process Development to Production
While we have described contamination as “almost impossible to locate, identify and
quantify” it does tend to obey its own “statistics” in most cases, moving within certain
ranges. The consequence of setting up a plasma process on a limited sample of parts
is understandably that the “sample average” will not equate to the “process average”.
Consequently, when moving from process development to production it is commonly
necessary to re‐position the plasma process to deal with the “worst case” that might be
found in production.
In addition, when introducing a plasma process one will, for the first time, be dealing
with a normalized, clean surface, a surface which essentially is known and “always the
same”. This almost always requires a re‐centering of the wire bond process.
Quality Management
It is always a nervous moment when discussing introducing a plasma process. There’s
always someone who wants to measure the surface before and after plasma to be sure
that it’s done what it should do. It can be done. In setting up the process a good
number of customers will try to understand why their wire bonding process is “out of
control”; what’s on the surface? X‐Ray photoelectron spectroscopy (XPS) alone or in
combination with Time‐of‐Flight Single Ion Mass Spectrometry (TOF‐SIMS) can
provide insights into what is on the surface. This information can also be useful in
deciding which plasma process to use. Neither technique, however, is practicable for
production QC and both are very expensive to operate. In addition, today’s
contaminant may not be tomorrow’s contaminant so any QC routine set up to monitor
contamination may end up measuring the wrong thing or miss something which has
appeared today, but wasn’t there yesterday. The approach which is followed in setting
up a plasma process nonetheless provides a level of security which in most cases is
adequate. The plasma process is set up to clean the bond pads in the “worst case”.
Without asking exactly what the contaminants are today, the plasma process is set up
to deliver a normalized, clean surface. This surface will deliver the wire bonding
statistics that were established during process development. Ultimately, all operators
end up using wire bonding statistics (which they were following anyway) as the
indicator that the plasma process is doing what it should do. Plasma systems are
designed to run a very repeatable process, with accurate control of gas flow, RF
power, process pressure, and process time. The experience in practice, with many
© Nordson MARCH
Page 8
www.NordsonMARCH.com
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
hundreds of machines running plasma before wire bonding, is that if the input to the
process remains unchanged (nature of semiconductor device and leadframe/substrate)
and the wire bonding process is stable, then the wire bonding statistics are very stable.
Compatibility and Concerns
Questions often arise about the effects of the plasma on the semiconductor devices;
are there issues with ESD, device parameter changes, charging, etc? Some of these,
such as overheating and sputtering with redeposition, are the possible consequences
of choosing incorrect process parameter settings and have been dealt with above. As a
reference point, it may be noted that virtually every microprocessor and every memory
device passes through a direct plasma prior to wire bonding (or in wafer level package
processing) without compromising function or reliability.
Used correctly, plasma is safe and effective. There are, nonetheless, classes of
semiconductor devices which are sensitive to direct plasmas. Devices with open
junctions, image sensors, EEPROMs and some types of power devices cannot be
exposed to direct plasma without causing performance changes or, in some cases,
catastrophic damage. In some cases, a plasma which essentially eliminates the RF
field, the presence of charged and energetic particles, and the light which is inherent in
the plasma process, can offer a solution. Nordson MARCH’s “ion‐free plasma” is one
such system.
It will always be the responsibility of a (potential) plasma user to ensure that there is no
compatibility issue with his devices, but your plasma equipment manufacturer can
assist in making this evaluation.
© Nordson MARCH
Page 9
www.NordsonMARCH.com
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
Figure 2. High Throughput Plasma
Treatment System (FlexTRAK)
Summary
The introduction of an appropriate plasma process prior to wire bonding will always
deliver a cleaner surface to bond to. Potential benefits are improved wire bond
statistics, improved device reliability, and the elimination of excursions caused by
non‐systemic influences (random pollution of the surfaces to be bonded by
uncontrolled factors).
John Maguire has been Business Manager, Europe for Nordson March since 2006.
With a degree in chemistry from the University of Bath (UK) and a doctorate in polymer
chemistry from the University of the South Bank (London, UK), Mr. Maguire combines
a strong technical background with thirty years of experience in the European PCB,
electronics, microelectronics and semiconductor industries. Working together with
specialized distributors in each region and segment, Mr. Maguire’s first strategic
objective is to ensure that plasma processing emerges from its „Black Box“ and takes
its place as a well known and well understood solution to many challenges in the
electronics industry and other market segments.
© Nordson MARCH
Page 10
www.NordsonMARCH.com
Plasma Clean to Reduce Wire Bond Failures
By John Maguire
Thank you for reading! Be sure to join us in social media for
the latest news and events.
Nordson MARCH
Phone: +1.925.827.1240
Email: [email protected]
2470-A Bates Avenue
Concord, CA USA 94520-1294
© Nordson MARCH
Page 11
www.NordsonMARCH.com